Pancreatic cancer is the fourth leading cause of cancer mortality in adults in the United States, with a 5- year survival rate that has remained at 1-3% for the past 25 years. At diagnosis, approximately 80% of pancreatic cancer patients have therapy-resistant locally advanced or metastatic disease with a median survival of less than 6 months. Even when stage I or II pancreatic cancer is apparently localized to the pancreas and surgically removed, 70% of those patients still develop liver metastases within 2 years after surgery. Current treatments of metastatic pancreatic cancer are largely ineffective. Pancreatic cancer thus poses one of the greatest challenges in cancer research. Although a genetic profile for pancreatic caner is emerging, the role of specific genetic alterations that initiate and mediate its cardinal clinical features of locally aggressive growth, metastasis, and chemotherapy resistance, remains unresolved. The underlying mechanisms, by which pancreatic cancer cells become invasive and metastatic, remain to be elucidated. Identification of the functions of the signature genetic and molecular alterations in pancreatic tumorigenesis and metastasis will provide a molecular basis for designing novel therapeutic approaches. Therefore, the long-term objective of the proposed research is to study the molecular basis of pancreatic metastasis using mouse models, which carry the signature genetic alterations found in this disease, and study the phenotypes induced by these genetic alterations. Our recent findings show that tropomyosin-related kinase BT1 (TrkBT1), which is a member of the protein tyrosine kinase receptor family, is overexpressed in metastatic pancreatic cancer cell line, Colo357L3.6pl, but not in the nonmetastatic parental cell line, Colo357FG. The overexpression of TrkBT1 has been related to occurrence of liver metastasis in human pancreatic adenocarcinoma. TrkBT1 was also the predominant overexpressed form of TrkB in other metastatic pancreatic cancer cell lines. Knock down of TrkBT1 by its shRNA induced anoikis. Overexpression of TrkBT1 in nonmetastatic Colo357FG cells induced colony formation in soft agar and liver metastasis in nude mice, but how the TrkBT1 variant, which lacks kinase-domain, induces pancreatic cancer metastasis remain to be further established. The hypothesis, the TrkBT1 variant play a key role in pancreatic cancer metastasis by activation of RhoA will be further tested. The specific aims are to determine whether the expression of GDI1, RhoA, and shRNA for GDI1 and RhoA will alter the metastatic potential of the pancreatic cancer cells, and to elucidate the mechanism by which the TrkBT1 induced the liver metastasis. A better understanding of the mechanisms of genetic alterations in induction of metastatic phenotypes will provide a basis for developing effective treatment strategies for pancreatic cancer. Targeting specific mutations and signaling pathways in pancreatic cancer may be one of the novel therapeutic interventions to treat cancer metastasis.